• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. Oct 2003; 185(19): 5735–5746.
PMCID: PMC193970

Genome-Wide Analyses Revealing a Signaling Network of the RcsC-YojN-RcsB Phosphorelay System in Escherichia coli

Abstract

In Escherichia coli, capsular colanic acid polysaccharide synthesis is regulated through the multistep RcsC→YojN→RcsB phosphorelay. By monitoring a hallmarked cps::lacZ reporter gene, we first searched for physiological stimuli that propagate the Rcs signaling system. The expression of cps::lacZ was activated when cells were grown at a low temperature (20°C) in the presence of glucose as a carbon source and in the presence of a relatively high concentration of external zinc (1 mM ZnCl2). In this Rcs signaling system, the rcsF gene product (a putative outer membrane-located lipoprotein) was also an essential signaling component. Based on the defined signaling pathway and physiological stimuli for the Rcs signaling system, we conducted genome-wide analyses with microarrays to clarify the Rcs transcriptome (i.e., Rcs regulon). Thirty-two genes were identified as putative Rcs regulon members; these genes included 15 new genes in addition to 17 of the previously described cps genes. Using a set of 37 two-component system mutants, we performed alternative genome-wide analyses. The results showed that the propagation of the zinc-responsive Rcs signaling system was largely dependent on another two-component system, PhoQ/P. Considering the fact that the PhoQ/P signaling system responds to external magnesium, we obtained evidence which supports the view that there is a signaling network that connects the Rcs system with the PhoQ/P system, which coordinately regulates extracellular polysaccharide synthesis in response to the external concentrations of divalent cations.

Escherichia coli and other enteric bacteria are capable of synthesizing an extracellular polysaccharide capsule, called colanic acid. A large number of genes are involved in capsular polysaccharide synthesis (36). The major gene cluster involved in colanic acid synthesis is located at 45.8 min on the canonical E. coli K-12 genetic map (38). The 5′-proximal gene cluster consists of the wza, wzb, and wzc genes, which are followed by the wca operon consisting of about 20 genes (http://www.genolist.pasteur.fr/Colibri/). The locus containing these genes was previously referred to as the cps operons (the capsular polysaccharide synthesis genes are referred to as cps genes for clarity below). The expression of these cps genes is coordinately regulated in response to a variety of environmental cues. Nevertheless, the physiological stimuli that induce capsular polysaccharide synthesis are not fully understood.

Extensive genetic studies have revealed that in E. coli there is a complex signaling circuitry that modulates expression of the cps genes (for reviews, see references 13 and 39). Together with the RcsA transcription factor and the Lon protease (10, 44), the RcsC-mediated histidine (His)→aspartate (Asp) phosphorelay system appears to play a major role in transcriptional regulation of the cps genes. By analogy with other His→Asp phosphorelay systems (1, 25), a unique model has been formulated for the multistep Rcs signaling system, in which the RcsC hybrid sensor His kinase senses certain environmental stimuli (42). The downstream signaling component is YojN (a histidine-containing phosphotransfer HPt factor), which serves as an intermediate for the phosphorelay. Eventually, the RcsB response regulator acquires a phosphoryl group from YojN, and the phosphorylated RcsB functions as a DNA-binding transcriptional regulator, which together with RcsA transcribes the cps genes. It has been suggested previously that another factor (RcsF) also plays a role in the Rcs signaling system (11). However, the importance of this putative outer membrane lipoprotein remains to be determined.

Interestingly, the Rcs signaling system is common in the sense that there are homologous systems not only in other enteric bacteria but also in nonenteric bacteria, including Salmonella enterica serovar Typhi, Vibrio cholerae, Klebsiella pneumoniae, Erwinia amylovora, Proteus mirabilis, and Pseudomonas aeruginosa. In S. enterica serovar Typhi, the Rcs signaling system modulates differential expression of an invasion protein (designated Sip), flagellin, and Vi antigen (2). The synthesis of certain extracellular polysaccharides in K. pneumoniae and E. amylovora is also regulated through the Rcs signaling systems (4, 28). The dimorphic and motile gram-negative bacterium P. mirabilis also uses the Rcs signaling system to regulate its characteristic swarming behavior (3, 15), like E. coli (17, 42). The recently released genome sequences of both V. cholerae and P. aeruginosa revealed that each of these bacteria has a homologous Rcs signaling system (18, 41). These findings indicate that the multistep RcsC→YojN→RcsB phosphorelay system is evolutionarily conserved in a wide variety of bacteria, in which this common adaptive response system is often associated with bacterial virulence and/or pathogenesis.

Needless to say, E. coli is the organism of choice for gaining general insight into the molecular mechanism underlying the common Rcs signaling systems. A thorough understanding of the E. coli Rcs signaling mechanism would answer the general question of why many virulent and/or pathogenic bacteria have the common Rcs signaling systems. To investigate this, we performed genome-wide analyses of the E. coli Rcs signaling system by using microarrays. We also employed a set of 37 two-component system mutants, in which each of the E. coli two-component systems was knocked-out one by one (32). The results of this study revealed not only a new view of the Rcs regulon but also an intriguing signaling linkage between the Rcs and PhoQ/P two-component systems in response to external divalent cations (zinc and magnesium).

MATERIALS AND METHODS

Bacterial strains and media.

E. coli K-12 strain ST001 and derivatives of this strain were the main organisms used. The ST001 strain [cps::lacZ thi ara Δ(pro-lac)] was constructed from CSH26 (19), as described previously (42). Another CSH26 derivative, strain GY37 [katE::lacZ thi ara Δ(pro-lac)], was also used (49), as was the MG1301 strain carrying the mgrB::lacZ gene on the chromosome (21). A set of 37 two-component mutants was constructed previously in the E. coli K-12 strain BW25113 background (32) (see Fig. Fig.6)6) by using the standard E. coli genetic procedures and/or the recently developed rapid procedures for constructing deletion mutants (7). The mutant alleles were each transferred into the ST001 background, one by one. In this study, a ΔrcsF derivative of ST001 was constructed by using the procedures used to construct deletion mutants (7). In this construct, the kanamycin resistance cassette replaced the entire rcsF gene. The primers used to construct an appropriate deletion cassette were 5′-CTCCTGATTCAATATTGACGTTTTGATCATACATTGAGGAAATACTATGGTGTAGGCTGGAGCTGCTTC and 5-GCCTATTTGCTCGAACTGGAAACTGCTCATTTCGCCGTAATGTTAAGCGATTCCGGGGATCCGTCGACC. The E. coli cells were grown mainly at either 37 or 20°C in EB medium, which contains 1.8% (wt/vol) Eiken's broth (Eiken Chemical Co. Ltd., Tokyo, Japan) and 0.4% glucose in 50 mM morpholinepropanesulfonic acid (MOPS) buffer (pH 7.0). When required, zinc, magnesium, and other divalent cations were added. Occasionally, Luria-Bertani medium was also used.

FIG. 6.
β-Galactosidase activity, determined by monitoring the cps::lacZ reporter gene in a set of 37 two-component system mutants. ST001 (wild type) and 37 two-component system mutants (each designated by a gene designation[s]) were grown in EB medium ...

Enzyme assay.

β-Galactosidase activity was measured by Miller's method, with slight modifications (23). The cells were grown as described above (except as noted below). They were suspended in 250 mM sodium phosphate (pH 7.1) for accurate measurement of cell density. A portion of a cell suspension was used for a β-galactosidase assay after permeabilization with toluene.

RNA isolation.

Cells were grown as described above or as described below. They were harvested at the mid-logarithmic growth phase. Total RNA was isolated with an RNeasy Maxi kit (Qiagen), as recommended by the supplier.

Preparation of E. coli DNA microarrays.

Custom glass slide microarrays (from Takara Shuzo, Kyoto, Japan) were spotted with 4,095 PCR products corresponding to full-length E. coli open reading frames, together with human β-actin genes as a negative reference. PCR products were generated by using a clone bank containing E. coli genes in the Archive vector as the template, as described previously (26). Other details were the same as those described previously (31, 32).

Preparation of labeled cDNA, hybridization, and data capture and analysis.

Appropriate pairs of RNA samples were prepared, and one sample was labeled with Cy3 and the other sample was labeled with Cy5. Each preparation was then tested twice by performing a microarray analysis. Thus, two values were obtained for each gene (or spot). The procedures used for hybridization, data capture, and data analysis were essentially the same as those described previously (31, 32). Briefly, by subtracting the local background value, we first corrected the intensity of each spot. In addition, a mean value for the intensity of the 24 negative control spots (human β-actin gene) was determined, together with the standard deviation (SD). Then spots were classified into three groups. For group 1 both the Cy3 signal intensity and the Cy5 signal intensity were greater than the mean plus one SD for the negative control. For group 2 either the Cy3 signal intensity or the Cy5 signal intensity was greater than the mean plus one SD for the negative control. For group 3 both the Cy3 signal intensity and the Cy5 signal intensity were less than the mean plus 1 SD for the negative control. Initially, we normalized spots classified in group 1 by defining the mean of the ratios (Cy5/Cy3) for all spots as 1.0. Group 2 spots with high Cy3 or Cy5 intensity values were selected. Group 3 spots were ignored as undetectable. Genes with significantly different expression were then selected by using the following criteria, provided that similar values were obtained in the two independent hybridizations. We basically selected genes that showed relative Cy5/Cy3 ratios that were <0.5 and genes that showed relative Cy5/Cy3 ratios that were >2.0 in both hybridizations as being down and up, respectively.

RESULTS

External stimuli that propagate the Rcs signaling system.

Except for osmotic shock (37), physiological stimuli that activate the Rcs signaling system have not been fully defined yet. We attempted to identify such external stimuli for the Rcs signaling system. An approximately 800-bp nucleotide sequence that contained the promoter of the colanic acid synthesis wza-wzb-wzc gene cluster was fused to the lacZ gene (40, 46). Then the fusion gene was introduced into the E. coli chromosome at the λ att site; the resulting strain was designated ST001, and the reporter gene was designated cps::lacZ (42). We then searched for growth conditions under which the cps::lacZ reporter gene was induced in a manner dependent on the Rcs signaling system. We found that when ST001 cells were grown at a low temperature (<20°C) in the presence of 0.4% glucose, the reporter gene was activated significantly (Fig. (Fig.1A1A and B). The basal medium contained 1.8% nutrient broth in 50 mM MOPS buffer (pH 7.0) (designated EB medium). When the cells were incubated at 20°C in EB medium containing 0.4% glucose, the level of β-galactosidase activity was significantly increased, while such induction of cps::lacZ was not seen in the absence of glucose (Fig. (Fig.1A)1A) or at 37°C (Fig. (Fig.1C1C and D). In these experiments, the cells were precultured at 37°C overnight. However, it should be noted that even if the cells were precultured at 20°C, essentially the same results were obtained. Thus, a cold shock appeared not to be implicated in this event.

FIG. 1.
β-Galactosidase activity, determined by monitoring the cps::lacZ reporter gene. (A) ST001 (wild-type) cells were grown overnight in EB medium at 37°C. Aliquots were inoculated into fresh EB medium with and without 0.4% glucose and incubated ...

The induction was dependent on the Rcs signaling components, RcsC and YojN (Fig. (Fig.1A).1A). As shown in Fig. Fig.1A,1A, the colonies on EB agar plates containing 0.4% glucose produced large amounts of extracellular polysaccharides at 20°C and thus exhibited mucous and glittering morphologies on the solid medium. This suggested that not only expression of the wza-wzb-wzc cluster genes but also expression of the whole system of colanic acid production was activated under these growth conditions. We also found that glucose was the most effective carbon source, and the other carbohydrates tested were not as effective (Fig. (Fig.1E1E).

External zinc is also an effective stimulus.

Although glucose serves as a stimulus for the Rcs signaling system, the observed β-galactosidase activity was only about 50 U, suggesting that the cps::lacZ reporter gene in ST001 may not be fully induced under these conditions. After extensively searching for other external stimuli, we found that addition of zinc (1 mM ZnCl2) to EB medium resulted in marked induction of the cps::lacZ reporter gene at 20°C in a manner dependent on the RcsC function (Fig. (Fig.2A2A and B). A lower concentration of zinc (0.5 mM) was also effective (data not shown), and these concentrations of zinc in EB medium were not deleterious for growth of cells (Fig. (Fig.2B).2B). In these experiments, the cells were precultured at 37°C overnight. However, it should be noted that even if the cells were precultured at 20°C, essentially the same results were obtained. Other divalent cations tested, including magnesium (30 mM MnCl2) and nickel (1 mM NiCl2), did not serve as stimuli (Fig. (Fig.2C2C and D). The effects of glucose and zinc on the cps::lacZ reporter gene were additive (Fig. (Fig.2A).2A). Thus, we found that the Rcs-mediated expression of cps::lacZ was fully induced under certain physiological growth conditions, namely, when cells were grown at a low temperature in the presence of glucose as a carbon source and in the presence of a relatively high concentration of external zinc. It should be noted that the levels of induction of cps::lacZ found in this study were much higher than the level of induction by an osmotic shock. As reported previously (42), the induction by an osmotic shock was less than 10 U under our experimental conditions.

FIG. 2.
β-Galactosidase activity, determined by monitoring the cps::lacZ reporter gene. (A) ST001 (wild type) and a derivative of this strain (ΔrcsC) were grown as described in the legend to Fig. Fig.1.1. In these experiments, glucose ...

RcsF is an essential component of the Rcs signaling system.

It was previously suggested that the rcsF gene product might be a component of the Rcs signaling system (11). RcsF was inferred to be an outer membrane lipoprotein. It was reported that when the rcsF gene on a multicopy plasmid was introduced into E. coli cells, the cps genes were markedly induced in a manner dependent on the Rcs signaling system, even at 37°C. In this study we confirmed this with ST001 (Fig. (Fig.3A).3A). Indeed, the expression of cps::lacZ was markedly elevated, provided that the rcsF gene on a multicopy plasmid was introduced into ST001. This event was completely dependent on each component of the Rcs signaling system (Fig. (Fig.3A).3A). Nevertheless, our results do not necessarily mean that RcsF is an essential component of the Rcs signaling system, because the multicopy rcsF gene was used in this experiment. Since we succeeded in identifying certain physiological growth conditions under which the cps genes were reproducibly induced, we were able to critically examine the longstanding question of whether RcsF is an essential component of the Rcs signaling system. A ΔrcsF derivative of ST001 was constructed, in which the entire coding region on the chromosome was deleted (see Materials and Methods). This derivative was then examined with regard to expression of cps::lacZ in response to external zinc at 20°C (Fig. (Fig.3B).3B). The results showed that the rcsF gene is an essential component of the Rcs signaling system, as far as signal transduction in response to glucose and zinc is concerned. Assuming that RcsF is an outer membrane protein, the following framework of the Rcs signaling system was formulated: RcsF (outer membrane protein)/RcsC (inner membrane protein) → YojN → RcsB/RcsA → Rcs regulon (target genes).

FIG. 3.
β-Galactosidase activity, determined by monitoring the cps::lacZ reporter gene. (A) ST001 (wild type [WT]) and mutant derivatives of this strain (ΔrcsC, ΔyojN, and ΔrcsB) were transformed with a multicopy plasmid harboring ...

Genome-wide microarray analyses with special reference to the Rcs signaling system.

Based on the framework and physiological stimuli that propagate the Rcs signaling system, we conducted genome-wide analyses with microarrays in order to characterize the Rcs regulon more extensively. Custom glass slide microarrays were spotted with PCR products corresponding to 4,095 full-length E. coli open reading frames (Takara Shuzo). This E. coli whole-genome microarray has been successfully employed in previous studies (31, 32). With this microarray system, systematic profile analysis of transcriptomes was carried out after one RNA sample was labeled with Cy3 and the other RNA sample was labeled with Cy5 (Fig. (Fig.4).4). First, the labeled transcripts were prepared from ST001 cells grown in glucose-containing EB medium at 20°C in the presence of zinc (1 mM), and they were compared with the transcripts from cells grown in the absence of zinc. This microarray analysis revealed the genes whose expression was induced by zinc (zinc-up genes). Second, the transcripts of ST001 (wild-type) cells grown at 20°C in the presence of glucose and zinc were compared with those of a mutant derivative of ST001 (ΔrcsC) grown under the same conditions. This microarray profile analysis revealed the genes whose induction was mediated by the RcsC function (ΔrcsC-down genes). To obtain reproducible data, a ΔrcsF mutant was also employed. The transcripts of ST001 (wild-type) cells grown at 20°C in the presence of glucose and zinc were compared with those of a mutant derivative of ST001 (ΔrcsF) grown under the same conditions. This microarray profile analysis revealed the genes whose induction was mediated by the RcsF function (ΔrcsF-down genes). When the data were considered together, we could systematically compile the genes whose expression was regulated through the RcsF/RcsC→YojN→RcsB signaling system in response to glucose and zinc.

FIG. 4.
Schematic representation of microarray data obtained with mutants with mutations in the Rcs signaling system. Three types of microarray analyses were carried out. First, the transcriptome of ST001 (wild-type) cells grown for 8 h at 20°C in EB ...

The results of the intensive microarray analyses are schematically summarized in Fig. Fig.4.4. First, 213 genes were identified as zinc-up genes. As expected, the cps::lacZ gene in ST001 was found in these genes (a critical internal reference for the reliability of the microarrays). Also, it was not surprising that such a large number of genes were affected by zinc, because when grown in the presence of external zinc, E. coli cells must change their physiologies in various ways. Second, 58 genes were identified as ΔrcsC-down genes, and 35 of these genes overlapped zinc-up genes (Fig. (Fig.4A).4A). These zinc-induced genes were assumed to be the genes whose expression is mediated by the RcsC function. Similarly, 120 genes were identified as ΔrcsF-down genes, and 49 of these genes overlapped zinc-up genes (Fig. (Fig.4B).4B). Then the 35 ΔrcsC-down and 49 ΔrcsF-down genes were compared with each other. Strikingly, 32 genes were found in both groups (Fig. (Fig.4C).4C). In addition to these genes, the cps::lacZ gene was recovered in this group, as mentioned above. Thus, we assumed that these 32 genes are most likely the genes whose expression is under control of the Rcs signaling system, directly or indirectly (these genes were tentatively referred to as Rcs regulon genes) (Fig. (Fig.4D4D).

It should be noted that 23 of the 58 ΔrcsC-down genes were not found in the zinc-up gene group. This fact may suggest that zinc and glucose regulate each set of genes through the Rcs signaling system. It may also be noteworthy that a much larger number of genes (120 genes) were affected in the ΔrcsF mutant than in the ΔrcsC mutant (58 genes). This fact may suggest that RcsF is also implicated in an as-yet-unknown regulatory system which is not dependent on the Rcs signaling system. In any event, these intriguing findings remain to be elucidated.

In addition to the cps genes, many other genes appear to be members of the Rcs regulon.

To examine the putative Rcs regulon genes more closely, the 32 genes discovered are listed in Fig. Fig.4D.4D. First, most of the known cps genes were identified as putative Rcs regulon members, as expected. These genes include the wzc gene, as well as the wca operon genes (46). Altogether, 17 of the cps genes were successfully identified as Rcs regulon members. In addition to these cps genes, the same microarray analyses revealed that 15 non-cps genes were putative Rcs regulon members. Some of these genes (eight genes) have previously been characterized to some extent, as noted in Fig. Fig.4D.4D. They included three of the osm genes (8, 12, 20, 29, 50). The bdm gene was characterized as the gene relevant for biofilm formation (33), whereas spy was shown to be a spheroplast-inducible gene (16, 34). The katE gene encodes an RpoS-dependent catalase (designated HPII) (43, 49).

As noted above, the 15 non-cps genes, including 7 uncharacterized y genes, seem to be Rcs regulon members. It may be noted that two of these genes, the bdm and osmC genes, are located next to each other on the chromosome but are oriented in opposite directions (Fig. (Fig.4).4). Also, the spy and ydjR genes appear to constitute an operon. As mentioned above, RcsB (together with RcsA) is a crucial DNA-binding regulator for the Rcs regulon. Previously, a consensus sequence (TAAGAATATTCCTA) was proposed as the RcsA/RcaB (heterodimer) recognition site (47). We searched for putative RcsA/RcsB recognition sites in the presumed promoter regions of the newly identified target genes. Many of these genes, if not all of them, have a short nucleotide sequence similar to the RcsA/RcsB recognition site in the appropriate promoter region (Fig. (Fig.5A).5A). However, some of the putative recognition sites are located far from the corresponding ATG initiation codons (e.g., katE, yajI, and yggE). Also, some of the promoters might also have an RcsB recognition site, which is independent of RcsA. Therefore, verification of RcsB and/or RcsA recognition sites in these putative Rcs regulon genes must await further inspection, which should include determination of each transcription start site. To address this issue further, an E. coli strain (designated GY37) carrying a katE::lacZ fusion gene on the chromosome was employed (49). The expression profiles of the katE::lacZ fusion gene were examined in terms of the Rcs signaling system (Fig. (Fig.5B).5B). Indeed, expression of katE::lacZ was induced in response to external glucose and zinc at 20°C. The induction of katE::lacZ was severely impaired in a set of rcs mutants.

FIG. 5.
Putative RcsA/RcsB recognition sequences in the Rcs regulon genes and β-galactosidase activity as determined by monitoring the katE::lacZ reporter gene. (A) Using the previously proposed RcsA/RcsB recognition consensus sequence (TAAGAATATTCCTA) ...

It should be emphasized that the proposed Rcs regulon genes are tentative, because in theory, the microarray analyses did not arrow us to discriminate whether a given gene that was identified is indeed the direct target of the transcriptional regulator in question. In this strict sense, some of the proposed Rcs regulon genes may or may not be direct targets of the Rcs signaling system. For example, the Rcs signaling system might somehow affect the activity of RpoS at the level of transcription, translation, or protein stability, which in turn could affect expression of the katE gene indirectly. With such limitations in mind, however, the microarray analyses performed in this study consistently revealed the set of 32 genes whose expression is under control of the Rcs signaling system directly or indirectly. In any case, the Rcs regulon genes, proposed above, appeared to be positively regulated by the Rcs signaling system. Theoretically, one can search for genes whose expression is negatively regulated by the Rcs signaling system. Such genes should be classified as zinc-down, ΔrcsC-up, and ΔrcsF-up genes. However, no such gene was revealed by our microarray data (data not shown). Thus, the Rcs signaling system acts mainly in positive regulation of the downstream target genes.

Another genome-wide analysis.

The genome-wide microarray analyses revealed a view of the Rcs regulon, as mentioned above. To further characterize the Rcs signaling system, we next took another genome-wide approach, as follows. We employed a set of 37 deletion mutants with mutations in the E. coli two-component systems (24, 25), in which each individual two-component gene was knocked out, one by one (32), as shown in Fig. Fig.6.6. In most of the mutants, an operon comprising a pair of sensor/regulator genes was deleted [for instance, Δ(envZ/ompR) and ΔphoQ/P]. In other instances, each corresponding sensor or regulator gene was deleted singly (for instance, ΔarcA). These 37 mutants were constructed previously, and they have been subjected to systematic analyses with microarrays previously (32). In this study, each of the mutant alleles was transferred into ST001 carrying the cps::lacZ reporter gene. For each derivative, expression of the cps::lacZ reporter gene was examined after cells were grown for 12 h at 20°C in EB medium containing glucose and zinc (Fig. (Fig.6).6). The results showed that as in the rcs mutants, the levels of induction of cps::lacZ were considerably reduced in the following genetic backgrounds: ΔarcA, Δ(envZ/ompR), and ΔphoQ/P. Rcs-mediated cps::lacZ expression was not significantly affected in the other genetic backgrounds. The results suggested that the Rcs signaling system might have linkage to some other phosphorelay systems.

To critically examine this possibility, the time course of cps::lacZ induction was monitored in each genetic background in question (Fig. (Fig.7).7). In both the ΔarcA and Δ(envZ/ompR) backgrounds (Fig. (Fig.7C7C and D), a slightly different cps::lacZ expression profile was obtained compared with that of the wild type (Fig. (Fig.7A).7A). In particular, the levels of expression at 12 h for ΔarcA (ca. 200 U) and Δ(envZ/ompR) (ca. 200 U) were significantly lower than the level of expression in the wild-type background (ca. 300 U). These observations were consistent with the findings shown in Fig. Fig.6,6, and thus the ArcA and/or EnvZ signaling system might have a link to the Rcs signaling system. However, the apparent effects of these mutants were subtle; thus, they were not characterized any further in this study. In contrast, the induction of cps::lacZ was more severely impaired in the ΔphoQ/P background (Fig. (Fig.7E).7E). Both the PhoQ sensor and PhoP regulator components were required for full induction of cps::lacZ in response to zinc (Fig. (Fig.7F7F and G). These results suggested that the PhoQ/P two-component genes might also be crucial determinants for regulating the cps genes in response to external zinc. In other words, there might be an intimate link between the Rcs and PhoQ/P signaling systems.

FIG. 7.
β-Galactosidase activity, determined by monitoring the cps::lacZ reporter gene in two-component mutant backgrounds. ST001 (wild type) and mutant derivatives of this strain were grown in EB medium containing 0.4% glucose and 1 mM ZnCl2 at 20°C, ...

Evidence for close linkage between the Rcs and PhoQ/P signaling systems.

We remembered the well-documented fact that the PhoQ/P signaling system functions in response to external magnesium (for a review, see reference 14). The phosphatase activity of PhoQ is modulated (or activated) through direct binding of magnesium ions to the PhoQ sensor, and thus the activity of the cognate PhoP response regulator is repressed at high concentrations of magnesium (45). The critical question then is, how do we explain the observed linkage between the PhoQ/P and Rcs signaling systems in response to zinc? We assumed that the PhoQ/P signaling system somehow functions upstream of the Rcs signaling system in response to zinc. If this is so, one can predict that the following events will occur: the cps::lacZ induction by zinc at 20°C in EB medium should be eliminated by the presence of a high concentration of magnesium in the medium, and the PhoQ/P signaling system should be attenuated under these conditions. The results of a critical examination of this hypothesis are shown in Fig. Fig.8A.8A. Indeed, induction of cps::lacZ by zinc (1 mM) was not observed, provided that magnesium (30 mM) was added concomitantly to the medium. This effect was dependent on the concentration of magnesium added (Fig. (Fig.8B8B).

FIG. 8.
β-Galactosidase activity, determined by monitoring the cps::lacZ reporter gene with cells grown in the presence of both magnesium and zinc. (A) ST001 cells were grown at 20°C in EB containing 0.4% glucose and 1 mM ZnCl2 with and without ...

This observation led us to envisage that the PhoQ sensor, not the RcsC sensor, might sense the external zinc signal. In other words, it was assumed that a high concentration of magnesium is a negative signal for the PhoQ sensor, whereas zinc might act as a positive signal. To test this idea, we examined expression of the mgrB gene, which is one of the hallmark targets of the PhoQ/P signaling system (21). When MG1301 cells carrying the mgrB::lacZ gene on the chromosome were grown at 20°C in EB medium with a high concentration of magnesium (30 mM MgCl2), the expression of mgrB::lacZ was severely repressed, as expected (Fig. (Fig.9A).9A). In magnesium-limited EB medium, however, a high, constitutive level of β-galactosidase activity was detected, also as expected (45). Interestingly, provided that zinc (1 mM) was added to the medium, the levels of β-galactosidase activity were further increased (Fig. (Fig.9A).9A). These results were best explained by assuming that the PhoQ/P signaling system responded not only to external magnesium but also to zinc, particularly when the concentration of external magnesium was relatively low. This event was confirmed by measuring expression of mgrB::lacZ with various concentrations of magnesium in the presence of zinc (Fig. (Fig.9B).9B). The results were completely consistent with the view that the PhoQ sensor is capable of responding to both magnesium and zinc in such a way that a high concentration of zinc stimulates the PhoQ function, while a high concentration of magnesium inhibits the PhoQ sensor.

FIG. 9.
β-Galactosidase activity, determined by monitoring the mgrB::lacZ reporter gene with cells grown in the presence of both magnesium and zinc. (A) MG1301 cells carrying the mgrB::lacZ reporter gene were grown at 20°C in EB medium containing ...

Critical genome-wide microarray analyses.

The results described above supported the presence of the plausible zinc→PhoQ/P signaling→Rcs signaling→Rcs regulon pathway, through which zinc propagates the Rcs system in a PhoQ/P-dependent manner. To critically examine this possibility, we conducted the following microarray analysis. Transcripts in ST001 (wild-type) cells grown at 20°C in the presence of glucose and zinc were compared with transcripts of a mutant derivative of ST001 (ΔphoQ/P) grown under the same conditions. This microarray profile analysis should have revealed the genes whose expression was mediated by the PhoQ/P function (ΔphoQ/P-down genes). A total of 100 genes were identified as such genes, and 61 of these genes overlapped zinc-up genes (Fig. 10A; also see Fig. Fig.4).4). These zinc-induced genes were assumed to be the genes whose expression is mediated by the PhoQ/P function. More importantly, it was found that 25 of the previously identified 32 Rcs regulon genes were included in the ΔphoQ/P-down group (Fig. 10B). In other words, only seven genes (e.g., wcaF and wcaK) had escaped the microarray analysis. When the intrinsic limitations of the microarray were considered, this result supported the presence of the zinc→PhoQ/P signaling→Rcs signaling→Rcs regulon signaling pathway.

FIG. 10.
Schematic representation of microarray data obtained with a mutant with a mutation in the PhoQ/P signaling system. Microarray analyses were carried out as indicted. The transcriptome of ΔphoQ/P cells grown at 20°C in EB medium containing ...

DISCUSSION

In this study, the Rcs phosphorelay signaling system in E. coli was characterized with a genome-wide viewpoint. Several intriguing views emerged from this analysis, as summarized in Fig. Fig.1111 (1). Not only the three Rcs phosphorelay factors but also RcsF was demonstrated to be a crucial component of the Rcs signaling system (Fig. (Fig.3)3) (2). The resulting RcsF/RcsC→YojN→RcsB/RcsA signaling system was propagated under certain physiological growth conditions, namely, when cells were grown at a low temperature in the presence of glucose as a carbon source and in response to a relatively high concentration of external zinc (Fig. (Fig.11 and and2)2) (3). By using microarray analyses, a view of the Rcs regulon was clarified by identifying a number of putative new target genes whose expression is positively regulated by the Rcs signaling system (Fig. (Fig.4).4). In addition to the well-known cps genes, a number of apparently unrelated genes, including osmB/C/Y, were suggested to be the members of Rcs regulon (Fig. (Fig.44 and and5)5) (4). The results of another genome-wide examination revealed a close linkage between the Rcs and PhoQ/P signaling systems, the latter of which is known to respond to external magnesium (Fig. (Fig.66 and and7).7). The results of this study supported the occurrence of the zinc (magneiusm)→PhoQ/Psignaling→Rcs signaling→Rcs regulon pathway (Fig. (Fig.88 to to10).10). Collectively, the results obtained in this study provided deeper insight into the E. coli phosphorelay systems, with special reference to the Rcs signaling system (Fig. (Fig.1111).

FIG. 11.
Proposed organization of the PhoQ/P-Rcs signaling network. The magnesium-regulated PhoQ/P signaling system appears to function upstream of the Rcs signaling system. The results of this study suggested that the PhoQ/P signaling system is activated by an ...

Microarray analysis is a powerful technique for systematically revealing the transcriptomes of organisms of interest. In E. coli, a large number of microarray data are accumulating rapidly (9, 30, 31, 32, 51). Previously, we have taken such genome-wide approaches to obtain a whole picture of the transcriptomes for the E. coli phosphorelay systems (32). To further extend our knowledge, in this study we used the same approach with special reference to the Rcs signaling system. The microarray analyses consistently revealed the members of the Rcs regulon. In addition to the known cps genes (17 genes), the results suggested that a number of new targets (15 genes) seemed to be activated through the Rcs signaling system. We need to confirm this assumption for each gene by another means, one by one. However, this view was based on the results of four independent microarray analyses (zinc-up, ΔrcsC-down, ΔrcsF-down, and ΔphoQ/P-down). Furthermore, we confirmed that the katE gene is regulated by the Rcs signaling system (Fig. (Fig.5),5), and the osmC gene has recently been reported to be under control of the Rcs signaling system (6, 8). In any case, the newly identified genes apparently are not relevant to colanic acid synthesis. Interestingly, however, many of them were predicted to encode cell envelope-associated proteins. These proteins include three putative lipoproteins (OsmB, YajI, and YggG), three putative periplasmic proteins (OsmY, Spy, and YggE), and two inner membrane proteins (OsmC and PhnQ). Thus, the Rcs signaling system may play more sophisticated roles in addition to its role in colanic acid synthesis. In fact, we previously showed that the Rcs signaling system is somehow involved in modulation of a characteristic behavior of E. coli cells during colony formation on the surfaces of agar plates, namely, swarming. It is also worth mentioning that the bdm gene is a biofilm-associated gene (33), and another gene (yggE) is an autoinducer-responsive gene implicated in quorum sensing (9). Thus, the Rcs signaling system might also be involved in characteristic E. coli physiological processes, such as swarming, biofilm formation, and/or quorum sensing.

The finding that there is intimate linkage between the Rcs and PhoQ/P signaling systems is intriguing since these signaling systems together respond to certain external divalent cations (zinc and magnesium) (Fig. (Fig.11).11). We do not know the molecular mode through which these two systems are linked. However, all of our results consistently supported the following presumed scheme: zinc and magnesium (stimuli) → PhoQ/P → X → RcsC/YojN/RcsB (where X is an unknown factor). Most likely, zinc directly signals PhoQ by modulating the kinase and/or phosphatase activities of PhoQ, and this zinc effect is counteracted by the magnesium effect, although the possibility that zinc directly modulates the RcsC sensor cannot be ruled out. It should also be noted that the ΔphoQ sensor mutant still responded to external zinc, albeit to a very low extent (Fig. (Fig.7G).7G). This might suggest that there is an as-yet-unknown senor signal for PhoR in response to zinc in the absence of the cognate PhoQ sensor. In any case, this residual activity was not affected by magnesium in the medium (data not shown). Also, expression of the rcsC, yojN, or rcsB gene might be dependent on the PhoQ/P functions. This is unlikely, because expression of these genes is not affected in the presence of the ΔphoQ/P mutation (Fig. (Fig.10).10). The proposed scheme shown in Fig. Fig.1111 is based on these intensive considerations.

Interestingly, this scheme led us to remember the well-documented PhoQ/P→[PmrD]→PmrB/A signaling network reported for an S. enterica serovar (for a review, see reference 14). In S. enterica, the PhoQ/P signaling system activates the PmrB/A signaling system during growth in the presence of a low magnesium concentration. In principle, the PmrB/A system is known to respond to external iron (5, 48). The unique gene product PmrD connects these two systems in such a way that the PhoQ/P signaling system regulates the production of PmrD, and the resulting protein activates the PmrB His kinase (22). Indeed, E. coli K-12 has an analogous set of genes, pmrD, pmrB (also called basS), and pmrA (or basR) (27). In the presumed zinc (stimulus)→PhoQ/P→X→RcsC/YojN/RcsB pathway, a hypothetical factor (X) might play a role similar to that of PmrD. Through both the PhoQ/P→Rcs and PhoQ/P→PmrB/A signaling networks, E. coli cells might be able to coordinately adapt to the states of external divalent cations, such as magnesium, iron, and zinc.

Many isolates of E. coli synthesize the slime polysaccharide colanic acid (references 35 and 36 and references therein). Small amounts of this capsule molecule are made by most E. coli strains, but certain growth conditions occasionally lead to dramatic increases in synthesis. Although the physiological role of colanic acid is still unclear, capsules are the outermost structures on an E. coli cell, and they may play a critical role in interactions between the bacterium and its environment by providing mechanisms for the cell to avoid host defenses in the intestinal tract. Capsules may also be important in the ability of E. coli to survive in extraorganismic environments. From the physiological point of view, the findings of this study showed that the environmental cation states, including the external levels of zinc and magnesium, appear to be critical environmental signals for capsular polysaccharide synthesis. The presumed framework shown in Fig. Fig.1111 provided several insight into the physiological role of the Rcs signaling system. It should be noted that the PhoQ/P signaling system regulates not only certain magnesium transporter genes but also genes involved in lipopolysaccharide modification (e.g., pagP) (14). Accordingly, the PhoQ/P→Rcs signaling network might allow E. coli cells to coordinately regulate cell surface polysaccharide (capsular polysaccharide and lipopolysaccharide) synthesis in response to the environmental states of divalent cations, such as magnesium and zinc. As elegantly proposed by Groisman (14) for the PhoQ/P→PmrB/A network in response to external magnesium and iron, the newly proposed PhoQ/P→Rcs network in response to external magnesium and zinc might also be important in natural habitats for maintenance of cell surface integrity and proper cell-cell interactions. These physiological processes might be crucial for swarming behavior on solid medium, quorum sensing, and/or biofilm formation, as mentioned above.

In summary, the genome-wide analyses performed in this study provided not only specific insight into the Rcs signaling system but also general insight into the E. coli phosphorelay systems. Since one can already see the impact of genome-wide analyses, such analyses for other phosphorelay systems should shed light on the global network of the His→Asp phosphorelay signal transduction systems in E. coli.

Acknowledgments

This study was supported by grants-in-aid from the Ministry of Education, Culture, Sports, Science and Technology of Japan (to T.M.) and from CREST of Japan Science and Technology (to H.M.).

We thank R. Utsumi for providing an E. coli strain (mgrB::lacZ).

REFERENCES

1. Appleby, J. L., J. S. Parkinson, and R. B. Bourret. 1996. Signal transduction via the multi-step phosphorelay: not necessarily a road less traveled. Cell 86:845-848. [PubMed]
2. Arricau, N., D. Hermant, H. Waxin, C. Ecobichon, P. S. Duffey, and M. Y. Popoff. 1998. The RcsB-RcsC regulatory system of Salmonella typhi differentially modulates the expression of invasion proteins, flagellin and Vi antigen in response to osmolarity. Mol. Microbiol. 29:835-850. [PubMed]
3. Belas, R., R. Schneider, and M. Melch. 1998. Characterization of Proteus mirabilis precocious swarming mutants: identification of rsbA, encoding a regulator of swarming behavior. J. Bacteriol. 180:6126-6139. [PMC free article] [PubMed]
4. Bereswill, S., and K. Geider. 1997. Characterization of the rcsB gene from Erwinia amylovora and its influence on exopolysaccharide synthesis and virulence of the fire blight pathogen. J. Bacteriol. 179:1354-1361. [PMC free article] [PubMed]
5. Chamnongpol, S., W. Dodson, M. J. Cromie, Z. L. Harris, and E. A. Groisman. 2002. Fe(III)-mediated cellular toxicity. Mol. Microbiol. 45:711-719. [PubMed]
6. Conter, A., R. Sturny, C. Gutierreze, and K. Cam. 2002. The RcsCB His-Asp phosphorelay system is essential to overcome chlorpromazine-induced stress in Escherichia coli. J. Bacteriol. 184:2850-2853. [PMC free article] [PubMed]
7. Datsenko, K. A., and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645. [PMC free article] [PubMed]
8. Davalos-Garcia, M., A. Conter, I. Toesca, C. Gutierrez, and K. Cam. 2001. Regulation of osmC gene expression by the two-component system rcsB-rcsC in Escherichia coli. J. Bacteriol. 183:5870-5876. [PMC free article] [PubMed]
9. DeLisa, M. P., C. F. Wu, L. Wang, J. J. Valdes, and W. E. Bentley. 2001. DNA microarray-based identification of genes controlled by autoinducer 2-stimulated quorum sensing in Escherichia coli. J. Bacteriol. 183:5239-5247. [PMC free article] [PubMed]
10. Ebel, W., and J. E. Trempy. 1999. Escherichia coli RcsA, a positive activator of colanic acid capsular polysaccharide synthesis, functions to activate its own expression. J. Bacteriol. 181:577-584. [PMC free article] [PubMed]
11. Gervais, F. G., and G. R. Drapeau. 1992. Identification, cloning, and characterization of rcsF, a new regulator gene for exopolysaccharide synthesis that suppresses the division mutation ftsZ84 in Escherichia coli K-12. J. Bacteriol. 174:8016-8022. [PMC free article] [PubMed]
12. Gordia, S., and C. Gutierrez. 1996. Growth-phase-dependent expression of the osmotically inducible gene osmC of Escherichia coli K-12. Mol. Microbiol. 19:729-736. [PubMed]
13. Gottesman, S., and V. Stout. 1991. Regulation of capsular polysaccharide synthesis in Escherichia coli K-12. Mol. Microbiol. 5:1599-1606. [PubMed]
14. Groisman, E. A. 2001. The pleiotropic two-component regulatory system PhoP-PhoQ. J. Bacteriol. 183:1835-1842. [PMC free article] [PubMed]
15. Gygi, G., M. Rahman, H. C. Lai, H. C., R. Carlson, J. Guard-Petter, and C. Hughes. 1995. A cell-surface polysaccharide that facilitates rapid population migration by differentiated swarm cells of Proteus mirabilis. Mol. Microbiol. 17:1167-1175. [PubMed]
16. Hagenmaier, S., Y. D. Stierhof, and U. Henning. 1997. A new periplasmic protein of Escherichia coli which is synthesized in spheroplasts but not in intact cells. J. Bacteriol. 179:2073-2076. [PMC free article] [PubMed]
17. Harshey, R. M., and T. Matsuyama. 1994. Dimorphic transition in Escherichia coli and Salmonella typhimurium: surface-induced differentiation into hyperflagellate swarmer cells. Proc. Natl. Acad. Sci. USA 91:8631-8635. [PMC free article] [PubMed]
18. Heidelberg, J. F., J. A. Eisen, W. C. Nelson, R. A. Clayton, M. L. Gwinn, R. J. Dodson, et al. 2000. DNA sequence of both chromosomes of the cholera pathogen Vibrio cholerae. Nature 406:477-483. [PubMed]
19. Ishige, K., S. Nagasawa, S. Tokishita, and T. Mizuno. 1994. A novel device of bacterial signal transducers. EMBO J. 13:5195-5202. [PMC free article] [PubMed]
20. Jung, J. U., C. Gutierrez, F. Martin, M. Ardourel, and M. Villarejo. 1990. Transcription of osmB, a gene encoding an Escherichia coli lipoprotein, is regulated by dual signals. J. Biol. Chem. 265:10574-10581. [PubMed]
21. Kato, A., H. Tanabe, and R. Utsumi. 1999. Molecular characterization of the PhoP-PhoQ two-component system in Escherichia coli K-12: identification of extracellular Mg2+-responsive promoters. J. Bacteriol. 181:5516-5520. [PMC free article] [PubMed]
22. Kox, L. F., M. M. Wosten, and E. A. Groisman. 2000. A small protein that mediates the activation of a two-component system by another two-component system. EMBO J. 19:1861-1872. [PMC free article] [PubMed]
23. Miller, J. H. 1972. Experiments in molecular genetics. Cold Spring Harbor Press, Cold Spring Harbor, N.Y.
24. Mizuno, T. 1997. Compilation of all genes encoding two-component phosphotransfer signal transducers in the genome of Escherichia coli. DNA Res. 4:161-168. [PubMed]
25. Mizuno, T. 1998. His-Asp phosphotranfer signal transduction. J. Biochem. (Tokyo) 123:555-563. [PubMed]
26. Mori, H., K. Isono, T. Horiuchi, and T. Miki. 2000. Functional genomics of Escherichia coli. Res. Microbiol. 151:121-128. [PubMed]
27. Nagasawa, S., K. Ishige, and T. Mizuno. 1993. Novel members of the two-component signal transduction genes in Escherichia coli. J. Biochem. (Tokyo) 114:350-357. [PubMed]
28. Nassif, X., N. Honore, T. Vasselon, S. T. Cole, and P. J. Sansonetti. 1989. Positive control of colanic acid synthesis in Escherichia coli by rmpA and rmpB, two virulence-plasmid genes of Klebsiella pneumoniae. Mol. Microbiol. 3:1349-1359. [PubMed]
29. Oh, J. T., Y. Cajal, M. E. Skowronska, S. Belkin, J. Chen, T. K. van Dyk, and M. K. Jain. 2000. Cationic peptide antimicrobials induce selective transcription of micF and osmY in Escherichia coli. Biochim. Biophys. Acta 15:43-54. [PubMed]
30. Oh, M. K., L. Rohlin, K. C. Kao, and J. C. Liao. 2001. Global expression profiling of acetate-grown Escherichia coli. J. Biol. Chem. 277:13175-13183. [PubMed]
31. Oshima, T., W. Wada, Y. Kawagoe, T. Ara, M. Maeda, Y. Masuda, et al. 2002. Genome wide analysis of deoxy-adenosine methltransferase-demidated control of gene expression in Escherichia coli. Mol. Microbiol. 45:673-695. [PubMed]
32. Oshima, T., H. Aiba, Y. Masuda, S. Kanaya, M. Sugiura, B. L. Wanner, et al. 2002. Transcriptome analysis of all two-component regulatory system mutants of Eschierichia coli K-12. Mol. Microbiol. 46:281-291. [PubMed]
33. Prigent-Combaret, C., O. Vidal, C. Dorel, and P. Lejeune. 1999. Abiotic surface sensing and biofilm-dependent regulation of gene expression in Escherichia coli. J. Bacteriol. 181:5993-6002. [PMC free article] [PubMed]
34. Raffa, R. G., and T. L. Raivio. 2002. A third envelope stress signal transduction pathway in Escherichia coli. Mol. Microbiol. 45:1599-1611. [PubMed]
35. Rick, P. D., and R. P. Silver. 1996. Enterobacterial common antigen and capsular polysaccharides: cellular and molecular biology, p. 104-122. In F. C. Neidhardt, R. Curtiss III, J. L. Ingraham, E. C. C. Lin, K. B. Low, B. Magasanik, W. S. Reznikoff, M. Riley, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella: cellular and molecular biology, 2nd ed. ASM Press, Washington, D.C.
36. Roberts, I. S. 1996. The biochemistry and genetics of capsular polysaccharide production in bacteria. Annu. Rev. Microbiol. 50:285-315. [PubMed]
37. Sledjeski, D. D., and S. Gottesman. 1996. Osmotic shock induction of capsule synthesis in Escherichia coli K-12. J. Bacteriol. 178:1204-1206. [PMC free article] [PubMed]
38. Stevenson, G., K. Andrianopoulos, M. Hobbs, and P. R. Reeves. 1996. Organization of the Escherichia coli K-12 gene cluster responsible for production of the extracellular polysaccharide colanic acid. J. Bacteriol. 178:4885-4893. [PMC free article] [PubMed]
39. Stout, V. 1994. Regulation of capsule synthesis includes interactions of the RcsC/RcsB regulatory pair. Res. Microbiol. 145:389-392. [PubMed]
40. Stout, V. 1996. Identification of the promoter region for the colanic acid polysaccharide biosynthetic genes in Escherichia coli K-12. J. Bacteriol. 178:4273-4280. [PMC free article] [PubMed]
41. Stover, C. K., X. Q. Pham, A. L. Erwin, S. D. Mizoguchi, P. Warrener, M. J. Hickey, et al. 2000. Complete genome sequence of Pseudomonas aeruginosa PAO1, an opportunistic pathogen. Nature 406:959-964. [PubMed]
42. Takeda, S., Y. Fujisawa, M. Matsubara, H. Aiba, and T. Mizuno. 2001. A novel feature of the multistep phosphorelay in Escherichia coli: a revised model of the RcsC→YojN→RcsB signaling pathway implicated in capsular synthesis and swarming behavior. Mol. Microbiol. 40:440-450. [PubMed]
43. Tanaka, K., K. Handel, P. C. Loewen, and H. Takahashi. 1997. Identification and analysis of the rpoS-dependent promoter of katE, encoding catalase HPII in Escherichia coli. Biochim. Biophys. Acta 30:161-166. [PubMed]
44. Trisler, P., and S. Gottesman. 1984. lon transcriptional regulation of genes necessary for capsular polysaccharide synthesis in Escherichia coli K-12. J. Bacteriol. 160:184-191. [PMC free article] [PubMed]
45. Vescovi, E. G., F. C. Soncini, and E. A. Groisman. 1996. Mg2+ as an extracellular signal: environmental regulation of Salmonella virulence. Cell 84:165-174. [PubMed]
46. Vincent, C., P. Doublet, C. Grangeasse, E. Vaganay, A. J. Cozzone, and B. Duclos. 1999. Cells of Escherichia coli contain a protein-tyrosine kinase, Wzc, and a phosphotyrosine-protein phosphatase, Wzb. J. Bacteriol. 181:3472-3477. [PMC free article] [PubMed]
47. Wehland, M., and F. Bernhard. 2000, The RcsAB box. Characterization of a new operator essential for the regulation of exopolysaccharide biosynthesis in enteric bacterial. J. Biol. Chem. 275:7013-7020. [PubMed]
48. Wosten, M. M., L. F. Kox, S. Chamnongpol, F. C. Soncini, and E. A. Groisman. 2000. A signal transduction system that responds to extracellular iron. Cell 103:113-125. [PubMed]
49. Yamashino, T., M. Isomura, C. Ueguchi, and T. Mizuno. 1998. The yhhP gene encoding a small ubiquitous protein is fundamental for normal cell growth of Escherichia coli. J. Bacteriol. 180:2257-2261. [PMC free article] [PubMed]
50. Yim, H. H., and M. Villarejo. 1995. osmY, a new hyperosmotically inducible gene, encodes a periplasmic protein in Escherichia coli. J. Bacteriol. 174:3637-3644. [PMC free article] [PubMed]
51. Zheng, M., X. Wang, L. J. Templeton, D. R. Smulski, R. A. LaRossa, and G. Storz. 2001. DNA microarray-mediated transcriptional profiling of the Escherichia coli response to hydrogen peroxide. J. Bacteriol. 183:4562-4570. [PMC free article] [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Compound
    Compound
    PubChem Compound links
  • Gene
    Gene
    Gene links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • MedGen
    MedGen
    Related information in MedGen
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • Protein
    Protein
    Published protein sequences
  • PubMed
    PubMed
    PubMed citations for these articles
  • Substance
    Substance
    PubChem Substance links
  • Taxonomy
    Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree